User-based topic, word, and sentiment analysis of Turkish tweets on platform X

Social media platforms, especially X (Twitter), provide rich data sources for understanding social and individual trends. While existing studies generally focus on general data sets, analyses at the individual user level remain limited. This study aims to fill this gap by presenting a web-based system that extracts and analyzes data from specific users from X. The developed system collects tweets from the desired user using the web scraping technique and preprocesses this data with steps specific to the Turkish language. Then, it applies three basic analyses with Latent Dirichlet Allocation (LDA): topic modeling, sentiment analysis, and word cloud generation. The system visualizes the results of topic distributions, sentiment graphs, and word clouds through a user-friendly interface. This study presents an original tool for understanding individuals' interests, emotional states, and mindsets in more detail by providing an in-depth user-based perspective.